CN112987003B - HFM signal separation method and system in active sonar - Google Patents
HFM signal separation method and system in active sonar Download PDFInfo
- Publication number
- CN112987003B CN112987003B CN202110159730.6A CN202110159730A CN112987003B CN 112987003 B CN112987003 B CN 112987003B CN 202110159730 A CN202110159730 A CN 202110159730A CN 112987003 B CN112987003 B CN 112987003B
- Authority
- CN
- China
- Prior art keywords
- signal
- time
- frequency
- transformation
- hfm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S15/00—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems
- G01S15/02—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems using reflection of acoustic waves
- G01S15/06—Systems determining the position data of a target
- G01S15/08—Systems for measuring distance only
- G01S15/32—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated
- G01S15/34—Systems for measuring distance only using transmission of continuous waves, whether amplitude-, frequency-, or phase-modulated, or unmodulated using transmission of continuous, frequency-modulated waves while heterodyning the received signal, or a signal derived therefrom, with a locally-generated signal related to the contemporaneously transmitted signal
Abstract
The invention relates to the technical field of underwater acoustic signal processing, in particular to a method and a system for separating HFM (high frequency modulation) signals in active sonar, wherein the method comprises the following steps: obtaining a transformation kernel function according to the related parameters of the used HFM signal; performing time-frequency transformation on the received signal by using a transformation kernel function; integrating the signals after time-frequency transformation to obtain incoherent accumulated output, determining time delay through a local output peak value, and performing band-pass filtering or band-stop filtering by taking the time delay as a center to obtain filtered signals; and performing inverse transformation on the filtered signal by adopting a rotation operator so as to filter out or filter out a corresponding signal component. In active sonar, the method can effectively detect HFM continuous wave signals, represents the corresponding relation between peak positions and instantaneous frequencies through time frequency, extracts and reconstructs target echo components by a narrow-band filtering method, and inhibits direct wave interference.
Description
Technical Field
The invention relates to the technical field of underwater acoustic signal processing, in particular to a method and a system for separating HFM (high frequency modulation) signals in active sonar.
Background
Most of the traditional Active sonars are Pulse Active Sonars (PAS). The method has the advantages of simple working mode and small signal processing calculation amount; but also has significant disadvantages such as high transmit power, short target illumination time, long probe update period, etc. In view of the above-mentioned problems, in recent years, continuous wave Active Sonar (CAS) technology has been considered important at home and abroad, and some progress has been made.
Compared with the traditional PAS, the CAS mode has the advantages of large processing gain, strong anti-interference capability, high target tracking and updating rate and the like. In contrast, CAS has higher requirements for the waveform design and processing method of the transmission signal: the method comprises the steps of firstly considering both the target updating rate and the time bandwidth gain, and secondly considering the multisource mutual interference problem of the CAS working in a multi-base mode. Under the influence of the field of Continuous Wave radar, when the Continuous Wave sounding technology is applied to the field of underwater sound sounding, a Linear Frequency Modulation Continuous Wave (LFMCW) signal is usually used as a transmitting signal. Subsequently, a Costas sequence, continuous single frequency signal, sinusoidal chirp, etc. waveform is applied to the CAS waveform design in sequence to increase the target update rate. Stefan M Murphy et al propose a sub-band filtering processing method of CAS, divide the linear frequency modulation signal into a plurality of sub-bands to carry out matching correlation processing, can improve the update rate, but correspondingly reduces the output signal-to-noise ratio. Liu Dali et al use the beat-fractional order Fourier algorithm for CAS detection to eliminate the LFMCW distance-velocity coupling, and obtain a signal processing gain equivalent to that of the matched filtering method, but do not consider the update rate problem. Zhou Jersen et al proposed a combined suppression method based on acoustic shielding and conventional directional nulling to eliminate direct wave interference of the CAS.
Hyperbolic Frequency-Modulated waves (HFM) is a doppler-tolerant signal, which is currently widely used in the field of underwater acoustic detection. Compared with LFMCW, the broadband Doppler insensitivity of the HFM continuous wave signal enables the HFM continuous wave signal to have unique advantages on detection of a moving target, doppler compensation is not needed in a sonar echo processing process, the operand is reduced, and the HFM continuous wave signal is a continuous wave signal suitable for detecting the underwater moving target. However, most of the existing signal processing methods of the CAS system are based on LFMCW, such as beat-fractional fourier transform, and are not completely suitable for HFM continuous wave signals.
For underwater sound long-time non-stationary signals, a single frequency domain or time domain analysis method can only obtain limited signal information, and the overall characteristics of the signals in the time domain and the frequency domain cannot be considered. The time-frequency analysis method can extract the instantaneous frequency, bandwidth, frequency delay and other time-frequency characteristics of the signals by constructing a time-frequency joint function and describing the strength and phase of the signals at different time and frequency simultaneously, can perform time-frequency filtering, and is a processing method suitable for underwater sound continuous wave signals. Traditional nonparametric time-frequency analysis methods include short-time fourier transform, wavelet transform, and wigner-well distribution, among others. However, the traditional time-frequency analysis method assumes that the signal is a local stationary signal to a certain extent, so that the analysis capability of the traditional time-frequency analysis method on the strong time-varying signal is weak. The parameterized time-frequency analysis constructs a corresponding transformation kernel according to the signal model, and can more accurately depict the local characteristics of the non-stationary signals. The transformation kernel of the parameterized time-frequency analysis directly determines the effect of the time-frequency analysis, so how to accurately design the transformation kernel is the key of the parameterized time-frequency analysis.
According to the hyperbolic frequency modulation CAS system in the bistatic mode, the HFM signal is periodically transmitted, the echo signal of the receiving end is analyzed, and target detection and positioning are carried out. During a sweep period, the transmit signal can be expressed as:
the instantaneous frequency of the transmitted signal can be expressed as:
where T denotes the period of the transmitted signal, A is the signal amplitude, f 0 As the time center frequency
T 0 For progressive time of signals
m is the signal frequency change rate m = f 0 /T 0 ，f max 、f min The maximum instantaneous frequency and the minimum instantaneous frequency of the transmitted signal respectively.
When there is a distance transmitting end R 1 Distance receiving end R 2 In a single emission cycleThe target echo signal may be expressed as:
wherein, K r Is a coefficient related to the reflection intensity and propagation loss of the target, eta is a time scale factor caused by Doppler effect, tau is echo signal time delay, and tau = (R) 1 +R 2 ) And c are sound velocities in water. So that the instantaneous frequency of the echo is
Since the HFM signal has Doppler invariance, under the conventional underwater sound detection scene, the Doppler compression effect of the HFM signal can be approximately considered to be equivalent to the shift tau of a frequency modulation function in time m So that the change law of the instantaneous frequency of the echo signal is constant, i.e.
Wherein
In bistatic mode, continuous direct wave components also exist in the received signal, and similar to equation (4), the instantaneous frequency of the direct wave in a single transmission period can be obtained as follows:
wherein tau is d ＝R 3 /c，R 3 Is the distance from the transmitting end to the receiving end.
Under a general CAS system, a time-frequency relationship diagram of an echo signal and a direct wave is shown in fig. 1. The echo often overlaps with the direct wave in time, and the echo analysis is easily affected by the direct wave.
Generalized Parameterized Time-Frequency analysis (GPTF Transform) is summarized and generalized by a number of Parameterized Time-Frequency analysis methods, which are defined as follows:
wherein
In the formula, k P Is a transformation kernel of the parameterized time-frequency analysis, P is a transformation kernel parameter,
for the frequency rotation operator defined by the parameter P,
is t 0 A frequency translation operator defined by a parameter P around the instant. When k is P (τ) =0, the above equation degenerates to a short-time fourier transform.
Without loss of generality, assume that the analytic signal of a certain signal is:
s(t)＝Aexp[j2π∫IF(t)dt] (9)
the instantaneous frequency of which is a function IF (t) of time. As shown in fig. 2, the parameterized time-frequency transform as defined in equation (10) is performed on s (t), and the process can be briefly summarized as follows: firstly, the signal is rotated in the time-frequency plane, i.e. the instantaneous frequency IF (tau) of the signal is subtracted by k P (τ); then making a translation transformation, i.e. increasing the instantaneous frequency by k P (t 0 ) (ii) a Final windowing function g σ And performing short-time Fourier transform on the signal. The frequency domain resolution of the parameterized time frequency transform result is determined by two parts: bandwidth of windowed signal portion Δ IF (t) 0 (ii) a σ) and a window function bandwidth of 1/σ. IF appropriate kernel functions and parameters P can be constructed, IF (tau) -k P (τ) is a constant at any time, Δ IF (t) 0 (ii) a Sigma) ≡ 0, the frequency domain resolution can be always the minimum value 1/sigma, and therefore the time-frequency representation with the highest energy concentration is obtained. Therefore, the frequency domain resolution of the parameterized time-frequency analysis is directly determined by the transformation kernel parameters, and the key for acquiring accurate time-frequency representation is to construct a transformation kernel k matched with the signal P (t)。
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a method and a system for separating HFM signals in active sonar.
In order to achieve the above object, the present invention provides an HFM signal separation method in active sonar, including:
obtaining a transformation kernel function according to the relevant parameters of the HFM signal;
performing time-frequency transformation on the received signal by using a transformation kernel function;
integrating the signals after time-frequency transformation to obtain incoherent accumulated output, determining time delay through a local output peak value, and performing band-pass filtering or band-stop filtering by taking the time delay as a center to obtain filtered signals;
and performing inverse transformation on the filtered signal by adopting a rotation operator, and separating out a corresponding signal component.
As an improvement of the above method, the transformation kernel function is obtained according to a related parameter of the HFM signal; the method specifically comprises the following steps:
the HFM signal s (t) is:
wherein A is a coefficient, T is a period, f 0 Is the time center frequency, T 0 Is a progressive time, t represents time;
from f 0 And T 0 Calculating the frequency change rate m = f of the HFM signal 0 /T 0 ；
The instantaneous frequency expression IF of the transmitted signal is obtained by s (t) is:
further obtain a transformation kernel function gamma Q (f) Comprises the following steps:
wherein f represents frequency.
As an improvement of the above method, the time-frequency transform is performed on the received signal by using a transform kernel function; the method specifically comprises the following steps:
wherein, G s And (t; Q) represents a signal after time-frequency transformation, t and Q respectively represent time and frequency after time-frequency transformation, S (theta) represents Fourier transformation of the HFM signal S (t), and theta is the angular frequency of a Fourier transformation domain.
As an improvement of the above method, the integration is performed on the signal after time-frequency transformation to obtain incoherent accumulated output, the time delay is determined by a local output peak value, and band-pass filtering or band-stop filtering is performed with the time delay as a center to obtain a filtered signal; the method comprises the following specific steps:
signal G after time-frequency conversion along frequency axis in rotating time-frequency domain s (t; Q) integrating to obtain incoherent accumulation output E (t) as:
wherein Q L And Q H A lower limit and an upper limit of the signal frequency Q, respectively;
for echo signal components in a single emission period, determining the time delay tau by local output peak as:
wherein, t 0 Indicating the starting moment of the transmission period, t 1 Indicating the end time of the transmission period;
performing band-pass filtering or band-stop filtering by taking the time delay tau as a center to obtain time-frequency representation of the filtered signal:
when performing band-pass filtering, the time-frequency representation of the filtered signal is:
when band-stop filtering is performed, the time-frequency representation of the filtered signal is as follows:
wherein, tau 0 Is the size of the time window defined according to the pulse width of the signal.
As an improvement of the above method, the filtered signal is subjected to inverse transformation by using a rotation operator, and a corresponding signal component is separated; the method specifically comprises the following steps:
applying corresponding rotation operators to the filtered signals
Inverse transformation is carried out to obtain a signal after inverse transformation
Reconstructing a signal
Thereby separating out the corresponding signal components.
An HFM signal separation system in active sonar comprises a transformation kernel function acquisition module, a time-frequency transformation module, a filtering module and a separation module; wherein, the first and the second end of the pipe are connected with each other,
the transformation kernel function acquisition module is used for obtaining a transformation kernel function according to the relevant parameters of the HFM signal;
the time-frequency transformation module is used for carrying out time-frequency transformation on the received signal by using a transformation kernel function;
the filtering module is used for integrating the signals after time-frequency conversion to obtain incoherent accumulated output, determining time delay through a local output peak value, and performing band-pass filtering or band-stop filtering by taking the time delay as a center to obtain filtered signals;
and the separation module is used for performing inverse transformation on the filtered signals by adopting a rotation operator to separate corresponding signal components.
As an improvement of the above system, the specific implementation process of the transformation kernel function obtaining module is as follows:
the HFM signal s (t) is:
wherein A is a coefficient, T is a period, f 0 Is the time center frequency, T 0 Is a progressive time, t represents time;
from f 0 And T 0 Calculating the frequency change rate m = f of the HFM signal 0 /T 0 ；
The instantaneous frequency expression IF of the transmitted signal is obtained by s (t) is:
further obtain a transformation kernel function gamma Q (f) Comprises the following steps:
wherein f represents a frequency.
As an improvement of the above system, the time-frequency transform module is implemented by the following steps:
wherein G is s And (t; Q) represents a signal after time-frequency transformation, t and Q respectively represent time and frequency after time-frequency transformation, S (theta) represents Fourier transformation of the HFM signal S (t), and theta is the angular frequency of a Fourier transformation domain.
As an improvement of the above system, the specific processing procedure of the filtering module is as follows:
signal G after time-frequency transformation along frequency axis in rotating time-frequency domain s (t; Q) integrating to obtain incoherent accumulation output E (t) as:
wherein Q L And Q H The lower limit and the upper limit of the signal frequency Q, respectively;
for echo signal components in a single emission period, determining the time delay tau by local output peak as:
performing band-pass filtering or band-stop filtering by taking the time delay tau as a center to obtain time-frequency representation of the filtered signal:
when band-pass filtering is performed, the time-frequency representation of the filtered signal is:
when band-stop filtering is performed, the time-frequency representation of the filtered signal is as follows:
wherein, tau 0 Is the size of the time window defined according to the pulse width of the signal.
As an improvement of the above system, the specific processing procedure of the separation module is as follows:
applying corresponding rotation operators to the filtered signals
Inverse transformation is carried out to obtain an inverse transformed signal
Reconstructing a signal
Thereby separating out the corresponding signal components.
Compared with the prior art, the invention has the advantages that:
1. in active sonar, the method can effectively detect HFM continuous wave signals, represents the corresponding relation between peak positions and instantaneous frequencies through time frequency, extracts and reconstructs target echo components by a narrow-band filtering method, and inhibits direct wave interference;
2. the parameterized time-frequency analysis method adopted by the invention is applied to the CAS detection field, can obtain better processing effect, realizes better estimation of echo time-frequency parameters, is beneficial to detection, positioning and characteristic analysis of targets, and can better realize signal detection and multi-target resolution according to the used HFM signal parameters.
Drawings
FIG. 1 is a schematic diagram of a hyperbolic frequency modulation CAS time-frequency relationship in the prior art;
FIG. 2 is a schematic diagram of a parametric time-frequency analysis principle;
FIG. 3 is a schematic diagram of a process flow of HFM continuous wave signals;
fig. 4 is a flow chart of the HFM signal separation method in the active sonar of the present invention;
FIG. 5 (a) is a time-frequency representation of the HFM signal of simulation example 1;
FIG. 5 (b) is an instantaneous frequency estimate of the HFM signal of simulation example 1;
FIG. 5 (c) is a HFM signal time-frequency representation accumulation output of simulation example 1;
fig. 6 is a receiver operating curve of simulation example 1;
FIG. 7 (a) is a received signal time-frequency representation of simulation example 1;
FIG. 7 (b) is a received signal rotation time-frequency representation of simulation example 1;
FIG. 7 (c) is a received signal time-frequency representation accumulation output of simulation example 1;
FIG. 8 (a) shows the extraction of echo components by narrow-band filtering of the received signal of simulation example 1
Fig. 8 (b) is a reconstructed target echo waveform of simulation example 1;
FIG. 9 is a received signal instantaneous frequency estimate of simulation example 1;
FIG. 10 (a) is a received signal time-frequency representation of simulation example 2;
FIG. 10 (b) is a received signal rotation time-frequency representation of simulation example 2;
FIG. 10 (c) is a received signal time-frequency representation accumulation output of simulation example 2;
FIG. 10 (d) shows the echo reconstruction result of simulation example 2
Fig. 10 (e) is a received signal instantaneous frequency estimation of simulation example 2;
fig. 10 (f) is a received signal positioning result of simulation example 2.
Detailed Description
The technical route of the invention is as follows: firstly, a signal processing method based on parameterized time-frequency analysis is provided for a CAS system with an HFM signal as a transmitting signal. According to parameters of a transmitting signal, a nonlinear transformation kernel corresponding to a hyperbolic function form is designed, a frequency domain signal is adopted to construct a time-frequency joint function based on HFM signal characteristics, and time-frequency representations of a plurality of echoes are obtained simultaneously. Secondly, based on the algorithm of parameterized time-frequency analysis, the echo and direct wave components in the multi-base sonar receiving signals can be separated, so that the direct wave interference is inhibited, and the detection performance is improved.
The algorithm of the invention is implemented by the following steps:
step 1) known HFM emission signal
Obtaining relevant parameters: emission signal period: t, time center frequency: f. of 0 The signal progressive time: t is a unit of 0 The signal frequency change rate: m = f 0 /T 0 。
Obtaining an instantaneous frequency representation of the transmitted signal:
thereby obtaining transformed kernel functions
Step 2) using the kernel function to perform time-frequency transformation on the received signal
Wherein S (θ) represents a fourier transform of the original signal; for a received signal containing a plurality of HFM continuous wave components, the transformed time-frequency distribution is a plurality of straight lines vertical to the delay axis, and the position on the delay axis represents the receiving delay tau of the component.
Step 3) firstly, integration is carried out along the frequency axis in the rotating time-frequency domain to obtain incoherent accumulated output,
wherein Q is L 、Q H Respectively, the upper and lower limits of the signal frequency.
For echo signal components within a single transmit period, the time delay tau is determined by the local output peak,
then, band-pass filtering and band-stop filtering are carried out by taking the time delay tau as the center, and then the time-frequency representation only containing the signal component can be obtained.
When the band-pass filtering is performed:
when the band-stop filtering is carried out:
wherein, tau 0 Is the size of the time window defined according to the pulse width of the signal.
And 4) performing inverse transformation on the filtered signal by adopting a corresponding rotation operator:
the signal components are reconstructed to separate out the corresponding signal components.
Generally, the direct wave interference can be suppressed by utilizing band-stop filtering; weak echo signals are extracted by means of band-pass filtering. When a plurality of underwater targets exist, the algorithm can separate echo signal components of different targets and acquire echo parameters one by one, so that a plurality of target echoes are analyzed.
The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and examples.
Example 1
The main flow of signal processing based on parameterized time-frequency analysis is shown in fig. 3. Firstly, band-pass filtering and beam forming are carried out on the received signals to obtain array gain, then, centralized time-frequency representation of the received signals is obtained through parameterized time-frequency analysis, and transformation kernel parameters of the centralized time-frequency representation can be determined by parameters of the transmitted signals. And setting a detection threshold to carry out peak value detection on the envelope peak value represented by the time frequency, and extracting time frequency characteristic parameters. And after estimated values such as time delay, time-frequency curves and the like are obtained, the target is positioned by combining a bistatic active detection positioning method.
Fig. 4 is a flowchart of the HFM signal separation method in active sonar according to the present invention, and the following describes a kernel function design method and a direct wave interference suppression method of an HFM time-frequency analysis method, respectively.
Designing a time-frequency transformation kernel function of hyperbolic frequency modulation:
in the design of the nonlinear transformation kernel function, a time-frequency characteristic approximation principle is usually used, a certain type of nonlinear transformation kernel in a general form is selected at first, and then the transformation kernel is used for carrying out multiple cyclic approximation refinement on a target signal, so that the most suitable transformation kernel parameter is obtained. The method needs to artificially select a fitting function, has large iterative calculation amount, cannot be directly used for multi-component signals, and has poor applicability to sonar detection. Based on the particularity of a sonar transceiving system and the time-frequency characteristic of HFM, a kernel function design method for a CAS system is provided below.
As can be seen from the formulas (5) and (6), the echo component and the direct wave component in the received signal both have the same instantaneous frequency variation law as the transmitted signal, and the instantaneous frequency functions differ by only one time delay, i.e. IF r (t)＝IF s (t-τ-τ m ). Aiming at the characteristics, the definition of generalized parameterized time-frequency analysis is converted into a frequency domain form according to the duality of the time-frequency domain, as follows:
wherein
In the formula, S (θ) represents fourier transform of an original signal;
and
respectively a frequency rotation operator and a translation operator; gamma ray Q And (omega) is a frequency domain transform kernel function. The parameterized time-frequency transformation principle of the frequency domain is completely the same as the time domain mode, and the difference is that the original signal mapped to the two-dimensional time-frequency surface by the parameterized time-frequency transformation principle is a one-dimensional frequency domain signal, and the corresponding operator is a function of frequency; whereas the latter is a one-dimensional time domain signal, the corresponding operator is a function of time.
Similarly, the key to the parameterized time-frequency analysis of the frequency domain is the kernel function γ Q And (4) determining. Suppose that the inverse of the instantaneous frequency function IF (t) of a signal (i.e. the local frequency delay function) is IF -1 (f) When the kernel function γ is Q (f)＝IF -1 (f)+τ c The time-frequency representation resolution can reach the minimum value, wherein tau c Is a time delay constant.
From equation (5), the inverse function of the instantaneous frequency function of the HFM received signal is:
therefore, only the kernel function is required
Can make the received signal at any time have
Namely, the parameterized time-frequency analysis result can obtain the minimum resolution.
The kernel function in the form of frequency domain is constructed by transmitting signal parameters, and the parameterized time-frequency analysis can simultaneously process HFM receiving signals containing a plurality of components (such as a plurality of target echoes or direct waves) to obtain the overall time-frequency representation of the receiving signals without performing operations such as successive filtering or hierarchical computation iteration. And after the time frequency representation with the highest energy concentration is obtained, performing peak detection along a time axis or a frequency axis in a time frequency domain to obtain ridge coordinates, and estimating various parameters of the received signals.
The HFM direct wave suppression method based on the parametric time-frequency analysis comprises the following steps:
in the CAS system, because of the bistatic mode, the energy of the direct wave is usually much higher than the echo component, and considering the influence of factors such as ocean channel multipath effect, the side lobe of the direct wave often masks the echo component, which makes it difficult to obtain the time-frequency parameter of the echo component in the time-frequency representation. Therefore, before positioning, the echo components need to be considered to be separated so as to eliminate the influence of the direct wave.
The basic idea of the method for separating the components of the received signal based on the parameterized time-frequency analysis is similar to that of the generalized demodulation time-frequency analysis method. By definition, in parametric time-frequency analysis, the rotation operator acts to rotate the time-frequency features of a signal over the time-frequency plane, while the translation operator acts to translate the energy of the signal to the location of the ridge of its true time-frequency feature. Thus, without the addition of the translation operator, the mathematical representation of the parameterized time-frequency analysis of the frequency domain degenerates to:
wherein the content of the first and second substances,
which is the frequency-domain signal after the rotation,
is a rotation operator determined by a kernel function. The above equation may be referred to as a parameterized rotated time-frequency transform. For a signal s (t) of the form of equation (9), if the window function is not considered, the transformation result can be expressed as:
inverse transform of the above formula
To make it
G s (t)＝δ(t-τ c )，
Namely, it is
S(θ)＝2πexp[j∫γ Q (θ)dθ+τ c ]，
Then the kernel function should be taken
γ Q (f)＝IF -1 (f)-τ c 。
I.e. a local frequency delay function of IF -1 (f)＝γ Q (f)+τ c The signal of (2), the time-frequency representation output peak value is concentrated on a straight line perpendicular to the delay axis, and the expression is t = τ c 。
Taking into account the HFM continuous wave received signal containing the direct wave component, the kernel function is taken when the rotation of the parameter is changed
The rotation time frequency is expressed as a plurality of straight lines perpendicular to the time delay axis, and the position on the time delay axis represents the receiving time delay tau of the component. For echo signal components in a single emission period, firstly, integration is carried out along a frequency axis in a rotating time-frequency domain to obtain incoherent accumulated output, a time delay tau is determined through an output peak value, and then narrow-band filtering is carried out by taking the time delay tau as a center, so that time-frequency representation only containing the signal components can be obtained. And then, performing inverse transformation as shown in the formula (17) on the filtered signal by adopting a corresponding rotation operator, and reconstructing the signal component, thereby separating an echo signal component and achieving the effect of inhibiting direct wave interference. And finally, carrying out parameterized time-frequency analysis on the separated echo signal components to obtain time-frequency representation of the echo, and extracting a time-frequency curve. When a plurality of underwater targets exist, the algorithm can also separate echo signal components of different targets and acquire echo parameters one by one, so that the targets are positioned.
Example 2
The embodiment 2 of the invention provides an HFM signal separation system in active sonar, which comprises a transformation kernel function acquisition module, a time-frequency transformation module, a filtering module and a separation module; wherein the content of the first and second substances,
the transformation kernel function acquisition module is used for obtaining a transformation kernel function according to the relevant parameters of the HFM signal;
the time-frequency transformation module is used for carrying out time-frequency transformation on the received signal by using a transformation kernel function;
the filtering module is used for integrating the signals after time-frequency conversion to obtain incoherent accumulated output, determining time delay through a local output peak value, and performing band-pass filtering or band-stop filtering by taking the time delay as a center to obtain filtered signals;
and the separation module is used for performing inverse transformation on the filtered signals by adopting a rotation operator to separate corresponding signal components.
The advantages of the present invention are further illustrated by a simulation example and an experimental data calculation example.
Simulation example 1, simulation data example analysis
(1) Parameterized time-frequency analysis performance simulation
According to the parameterized time-frequency analysis principle, white noise is uniformly distributed on a time-frequency plane, so that no aggregation peak occurs in time-frequency representation. Ideally, the theoretical processing gain of the parameterized time-frequency analysis on the HFM signal is:
where B is the signal bandwidth, Δ B σ Is the Gaussian window Bandwidth, Δ T, in time-frequency analysis σ Is the corresponding time domain length. In the formula, 10lg (2. Delta. B) σ *ΔT σ ) Time-frequency gain corresponding to the windowed portion, 10lg (B/Delta B) σ ) Corresponding to the non-coherent gain obtained by accumulation along the frequency axis.
In the simulation, the CAS transmission signal is assumed to be an HFM continuous wave signal with a frequency band of 400-500Hz and a period of 20s, and the time center frequency f thereof 0 =444Hz, rate of change of frequency m = 4.93(s) -2 ) The sampling rate was 2000Hz. The target is assumed to be a static target, the sonar transmitting end is 3km away from the target, the receiving end is 4.5km away from the target, the receiving signal-to-noise ratio is-3 dB, and direct wave influence does not exist. By the equation (14) and the transmitted signal parameters, the frequency domain transform kernel can be calculated
The simulated received signal was subjected to parametric time-frequency analysis with this transform kernel, with a gaussian window width of 20Hz, and the results are shown in fig. 5 (a). In the time-frequency representation result, the received signal energy is concentrated on a specific curve in a time-frequency domain, and the position of a ridge line formed by an envelope peak value corresponds to the time-frequency characteristic of the signal. In the time-frequency representation, the peak detection is performed on the frequency along the time axis, and the position of the ridge line is obtained, so as to obtain the estimated instantaneous frequency curve IF (t) of the signal, as shown in fig. 5 (b). Fig. 5 (c) shows the output obtained by performing non-coherent accumulation on the envelope peak of the time-frequency expression ridge line, the receiving delay estimation value corresponding to the peak is 4.98s, and the detection and positioning can be performed by combining the instantaneous frequency curve. The signal-to-noise ratios before and after analysis are shown in Table 1, and it can be seen that the statistical signal-to-noise ratio gain of the parameterized time-frequency analysis methodIn line with theoretical expectations.
TABLE 1 simulation signal GPTF processing method performance comparison result
| Type of signal | Receiving signal-to-noise ratio (dB) | Output signal-to-noise ratio (dB) | Simulation gain (dB) | Theoretical gain (dB) |
| HFM continuous wave | -3.0 | 25.5 | 28.5 | 29.0 |
On the basis of the simulation, a Monte Carlo statistical characteristic experiment method is used for analyzing the CAS detection performance and drawing a receiver working curve, as shown in FIG. 6. The detection probability curve of the parameterized time-frequency analysis method under the constant false alarm condition is given in the figure. From the simulation results it can be derived: when the false alarm probability is 0.1%, the detection probability of more than 99% can be achieved by parametric time-frequency analysis under the signal-to-noise ratio of-22 dB, and when the signal-to-noise ratio is reduced to-24 dB, the detection probability can still reach more than 80%; when the false alarm probability is 0.01%, the detection probability of 97% can still be achieved when the signal-to-noise ratio is-22 dB. The effectiveness of the parameterized time-frequency analysis method in CAS signal detection is verified.
(2) Direct wave interference suppression process simulation
In order to verify the effectiveness of the parameterized time-frequency analysis method on active sonar direct wave interference suppression, direct wave interference is added in CAS simulation conditions. The distance between a sonar transmitting end and a receiving end is assumed to be 6km, and HFM signal parameters and other environmental conditions are unchanged.
Fig. 7 (a) shows the result of parameterized time-frequency analysis of the received signal. The time-frequency representation of the method comprises two peak value crest lines of a direct wave component and an echo component, which are respectively corresponding to instantaneous frequency curves of the two components. Under simulation conditions, the signal-to-interference ratio of the target echo to the direct wave interference is close to-27 dB, and the instantaneous frequency of the echo is difficult to extract in the time-frequency representation. Fig. 7 (b) shows the result of a parameterized rotated time-frequency transform of a received signal with the same transform kernel, with the ridges corresponding to the signal components becoming straight lines perpendicular to the time axis. Fig. 7 (c) shows the result of incoherent gain obtained by integrating the time-frequency representation along the frequency axis in the rotating time-frequency domain, and the reception time delay of the direct wave and the target echo component can be obtained through the peak values of the incoherent gain, which are 3.95s and 4.98s respectively.
In the rotation time-frequency representation, band-pass filtering is performed around 4.98s, and then a target echo waveform is reconstructed through rotation time-frequency inverse transformation, as shown in fig. 8 (a) and (b). The echo signals are individually subjected to parameterized time-frequency analysis to obtain time-frequency representation, and a complete instantaneous frequency curve of the received signals can be obtained by combining the time-frequency representation of the separated direct wavefront, as shown in fig. 9. The simulation result verifies the effectiveness of the algorithm on the direct wave interference suppression, and target echo and direct wave components can be effectively separated through the operation in a signal time-frequency domain, so that the target information can be conveniently extracted. The parameterized time-frequency analysis method does not relate to receiving directivity, so that the method can be combined with the traditional spatial filtering method to further inhibit direct wave interference.
Simulation example 2, experimental data example analysis
In 2018, in 4 months, a shallow sea active sonar detection test is performed in the south sea, a bistatic active sonar mode is adopted in the test, the 'experiment No. 2' is used as a transmitting ship, a transmitting sound source is a hanging and placing transducer, and a transmitting signal is a HFM long signal with the frequency band of 400-500Hz and the pulse width of 20 s; the detection target is an underwater hull part of the catamaran of experiment No. 1; and a seabed horizontal receiving array is used as a receiving end. The experimental process is divided into 4 stations, and each station transmits 4 sets of HFM continuous wave signals. In the experiment, the launching ship is stopped and drifted to reduce the influence of self-noise; the target ship sails along a preset air route and can be regarded as a low-speed moving target; the receiving end interference is mainly environmental noise and high-intensity direct waves.
Firstly, a target azimuth is obtained through array processing, and then parameterized time-frequency analysis processing is carried out on a receiving signal formed by a conventional wave beam. The frequency width of the time-frequency analysis window is 20Hz, the length of a processing signal segment is 40s, and the processing signal segment comprises an echo wave and a direct wave in the same transmission period. And (3) carrying out parameterized time-frequency analysis processing on each section of received signals, comparing signal-to-noise ratio gains before and after the processing, and taking the average value of each station sample, wherein the result is shown in a table 2.
TABLE 2 Experimental signal gain comparison results (average for each station)
| Standing position | Experimental gain (dB) | Theoretical gain (dB) |
| 1 | 28.3 | 29.0 |
| 2 | 28.4 | 29.0 |
| 3 | 26.4 | 29.0 |
| 4 | 28.9 | 29.0 |
Fig. 10 (a) shows the parameterized time-frequency analysis of a segment of the received signal at a transmitting vessel 5.175km from the target vessel, with time 0 corresponding to the time of signal transmission. It can be seen from the time-frequency representation that the target echo of the signal of this segment is very close to the receiving time of the direct wave, and the masking range caused by the direct wave is large. Fig. 10 (b) and (c) show the corresponding rotation time-frequency transformation result and the time-frequency domain accumulation output, respectively. In the output result, the peak value of the echo signal is positioned at a direct wave side lobe, and the corresponding receiving time delay is 7.9s. The echo waveform reconstructed by the time-frequency domain filtering and the inverse transform is shown in fig. 10 (d). Using the reconstructed target echo, an instantaneous frequency curve of the target echo is obtained by time-frequency analysis, as shown in fig. 10 (e). When combined with the frequency analysis, array processing results and bistatic sonar system positioning method, the target positioning at the corresponding moment can be given, as shown in fig. 10 (f). Wherein the relative coordinates of the transmitting vessel and the receiving array are obtained from the GPS data.
The experimental result shows that, on one hand, the parameterized time-frequency method can effectively detect HFM continuous waves, obtain time-frequency processing gain close to a theoretical value and simplify the positioning processing flow; on the other hand, the parameterized time-frequency method can separate direct wave and target echo components through filtering and reconstruction in a time-frequency domain, and effectively inhibit direct wave interference in the multi-base CAS.
The CAS system HFM signal processing method based on the parameterized time-frequency analysis can effectively detect HFM continuous wave signals, represents the corresponding relation between peak positions and instantaneous frequencies through the time frequency, extracts and reconstructs target echo components by a narrow-band filtering method, and inhibits direct wave interference. Simulation and experimental results show that the parameterized time-frequency analysis method is applied to the CAS detection field, can obtain better processing effect, realizes better estimation of echo time-frequency parameters, and is beneficial to detection, positioning and characteristic analysis of targets.
Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention and are not limited. Although the present invention has been described in detail with reference to the embodiments, it should be understood by those skilled in the art that the technical solutions of the present invention may be modified or substituted with equivalents without departing from the spirit and scope of the technical solutions of the present invention, and all of them should be covered by the scope of the claims of the present invention.
Claims (8)
1. A method of HFM signal separation in active sonar, the method comprising:
obtaining a transformation kernel function according to the relevant parameters of the HFM signal;
performing time-frequency transformation on the received signal by using a transformation kernel function;
integrating the signals after time-frequency transformation to obtain incoherent accumulated output, determining time delay through a local output peak value, and performing band-pass filtering or band-stop filtering by taking the time delay as a center to obtain filtered signals;
performing inverse transformation on the filtered signals by adopting a rotation operator, and separating corresponding signal components;
obtaining a transformation kernel function according to the relevant parameters of the HFM signal; the method specifically comprises the following steps:
the HFM signal s (t) is:
wherein A is a coefficient, T is a period, f 0 Is the time center frequency, T 0 T represents time, which is progressive time;
from f to f 0 And T 0 Calculating the frequency change rate m = f of the HFM signal 0 /T 0 ；
The instantaneous frequency expression IF of the transmitted signal is obtained by s (t) is:
further obtain a transformation kernel function gamma Q (f) Comprises the following steps:
wherein f represents a frequency.
2. The HFM signal separation method in active sonar according to claim 1, wherein the time-frequency transform is performed on the received signal using a transform kernel function; the method specifically comprises the following steps:
wherein, G s (t; Q) represents the signal after time-frequency transformation, t and Q represent the time and frequency, respectively, after time-frequency transformation, S (theta) represents the Fourier transform of the HFM signal S (t), and theta is the angular frequency of the Fourier transform domain.
3. The method for separating the HFM signal in the active sonar according to claim 2, wherein the integration is performed on the signal after the time-frequency transformation to obtain a non-coherent accumulation output, a time delay is determined by a local output peak value, and band-pass filtering or band-stop filtering is performed with the time delay as a center to obtain a filtered signal; the method specifically comprises the following steps:
signal G after time-frequency transformation along frequency axis in rotating time-frequency domain s (t; Q) integrating to obtain incoherent accumulation output E (t) as:
wherein Q L And Q H A lower limit and an upper limit of the signal frequency Q, respectively;
for echo signal components in a single transmission period, determining the time delay tau by the local output peak as:
wherein, t 0 Indicating the starting moment of the transmission period, t 1 Indicating the end time of the transmission period;
performing band-pass filtering or band-stop filtering by taking the time delay tau as a center to obtain time-frequency representation of the filtered signal:
when performing band-pass filtering, the time-frequency representation of the filtered signal is:
when band-stop filtering is performed, the time-frequency representation of the filtered signal is as follows:
wherein, tau 0 Is the size of the time window defined according to the pulse width of the signal.
4. The HFM signal separation method in active sonar according to claim 3, wherein the filtered signal is inverse transformed using a rotation operator to separate out corresponding signal components; the method specifically comprises the following steps:
using corresponding rotation operators to filter the signal
Inverse transformation is carried out to obtain an inverse transformed signal
Reconstructing a signal
Thereby separating out the corresponding signal components.
5. An HFM signal separation system in active sonar is characterized by comprising a transformation kernel function acquisition module, a time-frequency transformation module, a filtering module and a separation module; wherein the content of the first and second substances,
the transformation kernel function acquisition module is used for obtaining a transformation kernel function according to the relevant parameters of the HFM signal;
the time-frequency transformation module is used for carrying out time-frequency transformation on the received signal by using a transformation kernel function;
the filtering module is used for integrating the signals after time-frequency conversion to obtain incoherent accumulated output, determining time delay through a local output peak value, and performing band-pass filtering or band-stop filtering by taking the time delay as a center to obtain filtered signals;
the separation module is used for performing inverse transformation on the filtered signal by adopting a rotation operator to separate out a corresponding signal component;
the specific implementation process of the transformation kernel function acquisition module is as follows:
the HFM signal s (t) is:
wherein A is a coefficient, T is a period, f 0 Is the time center frequency, T 0 Is a progressive time, t represents time;
from f 0 And T 0 Calculating the frequency change rate m = f of the HFM signal 0 /T 0 ；
The instantaneous frequency expression IF of the transmitted signal is obtained by s (t) is:
further obtain a transformation kernel function gamma Q (f) Comprises the following steps:
wherein f represents frequency.
6. The HFM signal separation system in active sonar according to claim 5, wherein the time-frequency transform module is implemented by:
wherein G is s (t; Q) represents the signal after time-frequency transformation, t and Q represent the time and frequency, respectively, after time-frequency transformation, S (theta) represents the Fourier transform of the HFM signal S (t), and theta is the angular frequency of the Fourier transform domain.
7. The HFM signal separation system in active sonar according to claim 6, wherein the specific processing procedure of the filtering module is:
signal G after time-frequency conversion along frequency axis in rotating time-frequency domain s (t; Q) integrating to obtain incoherent accumulation output E (t) as:
wherein Q is L And Q H A lower limit and an upper limit of the signal frequency Q, respectively;
for echo signal components in a single transmission period, determining the time delay tau by the local output peak as:
performing band-pass filtering or band-stop filtering by taking the time delay tau as a center to obtain time-frequency representation of the filtered signal:
when band-pass filtering is performed, the time-frequency representation of the filtered signal is:
when band-stop filtering is performed, the time-frequency representation of the filtered signal is as follows:
wherein, tau 0 Is the size of the time window defined according to the pulse width of the signal.
8. The HFM signal separation system in active sonar according to claim 7, wherein the specific processing procedure of the separation module is:
using corresponding rotation operators to filter the signal
Inverse transformation is carried out to obtain an inverse transformed signal
Reconstructing a signal
Thereby separating out the corresponding signal components.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110159730.6A CN112987003B (en) | 2021-02-05 | 2021-02-05 | HFM signal separation method and system in active sonar |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110159730.6A CN112987003B (en) | 2021-02-05 | 2021-02-05 | HFM signal separation method and system in active sonar |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112987003A CN112987003A (en) | 2021-06-18 |
| CN112987003B true CN112987003B (en) | 2022-12-06 |
Family
ID=76347467
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110159730.6A Active CN112987003B (en) | 2021-02-05 | 2021-02-05 | HFM signal separation method and system in active sonar |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112987003B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115022145B (en) * | 2022-07-01 | 2023-09-29 | 舟山美通信息技术有限责任公司 | LFM signal and HFM signal distinguishing method based on delay autocorrelation |
| CN116973901A (en) * | 2023-09-14 | 2023-10-31 | 海底鹰深海科技股份有限公司 | Algorithm application of time-frequency analysis in sonar signal processing |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6130857A (en) * | 1985-03-15 | 2000-10-10 | The United States Of America As Represented By The Secretary Of The Navy | Aural enhancement of active sonar systems |
| FR2853417A1 (en) * | 2003-04-01 | 2004-10-08 | Thales Sa | SIGNAL PROCESSING METHOD, AND ACTIVE SONAR USING THE SAME |
| CN103020479A (en) * | 2012-12-28 | 2013-04-03 | 上海交通大学 | Signal instantaneous frequency estimation method based on nonlinear frequency modulation wavelet transformation |
| CN106931973A (en) * | 2017-03-14 | 2017-07-07 | 杭州电子科技大学 | High accuracy indoor locating system and method based on nonlinear FM pulse signal |
| CN112014813A (en) * | 2020-09-03 | 2020-12-01 | 中国人民解放军海军航空大学 | Sky-wave radar ionosphere pollution correction method and system |
-
2021
- 2021-02-05 CN CN202110159730.6A patent/CN112987003B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6130857A (en) * | 1985-03-15 | 2000-10-10 | The United States Of America As Represented By The Secretary Of The Navy | Aural enhancement of active sonar systems |
| FR2853417A1 (en) * | 2003-04-01 | 2004-10-08 | Thales Sa | SIGNAL PROCESSING METHOD, AND ACTIVE SONAR USING THE SAME |
| CN103020479A (en) * | 2012-12-28 | 2013-04-03 | 上海交通大学 | Signal instantaneous frequency estimation method based on nonlinear frequency modulation wavelet transformation |
| CN106931973A (en) * | 2017-03-14 | 2017-07-07 | 杭州电子科技大学 | High accuracy indoor locating system and method based on nonlinear FM pulse signal |
| CN112014813A (en) * | 2020-09-03 | 2020-12-01 | 中国人民解放军海军航空大学 | Sky-wave radar ionosphere pollution correction method and system |
Non-Patent Citations (2)
| Title |
|---|
| Improved instantaneous frequency estimation using an adaptive short-time Fourier transform;H.K. Kwok 等;《IEEE Transactions on Signal Processing》;20001231;第48卷(第10期);全文 * |
| 基于广义参数化时频变换的天波雷达电离层污染校正;薛永华 等;《信号处理》;20201231;第36卷(第12期);第2024-2031页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112987003A (en) | 2021-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112987003B (en) | HFM signal separation method and system in active sonar | |
| US9696423B2 (en) | Method for improving performance of a Sodar system | |
| CN107272005B (en) | Active positioning method based on target echo arrival time delay and arrival angle under reliable acoustic path | |
| CN111142105A (en) | ISAR imaging method for complex moving target | |
| CN103353591A (en) | Bistatic radar localization dimension reduction clutter suppression method based on MIMO | |
| CN113281727B (en) | Output enhanced beam forming method and system based on horizontal line array | |
| CN110346798B (en) | Bistatic synthetic aperture radar wavenumber domain efficient imaging processing method | |
| CN103353592A (en) | Bistatic radar multichannel combination dimension reduction clutter suppression method based on MIMO | |
| CN104360251A (en) | Ultrasonic signal time delay estimation method for partial discharging of potential transformer | |
| CN110763327A (en) | Single-vector hydrophone signal detection method under condition of plane baffle | |
| CN110632573B (en) | Airborne broadband radar space-time two-dimensional keystone transformation method | |
| CN112924959B (en) | Bandwidth synthesis-based distributed high-frequency ground wave radar target positioning and speed measuring method | |
| CN110109092B (en) | Radar speed measurement method based on time reversal in multipath environment | |
| CN115242583B (en) | Channel impulse response passive estimation method based on horizontal linear array | |
| CN104793197A (en) | Direct-wave suppression method based on IFFT frequency spectrum division method and gradient adaptive lattice filter | |
| CN113567978B (en) | Multi-base distributed radar collaborative imaging method | |
| CN115656994A (en) | Real-time calibration method for double-base active detection towed array formation | |
| CN112824927B (en) | Sky wave over-the-horizon radar ionospheric phase pollution correction method based on sparse filtering | |
| CN111077515B (en) | Target detection method based on analog television external radiation source radar | |
| CN103584886B (en) | Self-adaption apodization method based on phase coherent information | |
| CN113126029A (en) | Multi-sensor pulse sound source positioning method suitable for deep sea reliable acoustic path environment | |
| Sakhnini et al. | An experimental evaluation of robust near-field radar localization using a massive MIMO testbed | |
| CN117310671B (en) | Shallow sea sound source distance environment self-adaptive estimation method applying frequency dispersion elimination transformation | |
| CN112466330B (en) | Sound source level estimation method for noise source under multi-path channel | |
| Yan et al. | Low-sidelobe Imaging Method Utilizing Improved Spatially Variant Apodization for Forward-Looking Sonar |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |